writing a lot of code allows many experiments to be performed. this is valuable when the problem is not well-defined, or when solutions can't be evaluated unti

writing a lot of code allows many experiments to be performed. this is valuable when the problem is not well-defined, or when solutions can't be evaluated until tested on actual users or under partly-unknown real-world conditions. otherwise he's right, a problem should be studied carefully before any coding, for best efficiency